The Research Desk.

The most upvoted and starred AI research crossing the community today.

Last Brew Time: Oct 21, 2026, 12:12 PM PT

X.com Research Buzz

NLP
The Algorithmic Self-Portrait: Deconstructing Memory in ChatGPT
X.com
7016

The Algorithmic Self-Portrait: Deconstructing Memory in ChatGPT

Abhisek Dash, Soumi Das, Elisabeth Kirsten, Qinyuan Wu, Sai Keerthana Karnam, Krishna P. Gummadi, Thorsten Holz, Muhammad Bilal Zafar, Savvas Zannettou

Max Planck Institute for Software Systems, Ruhr University Bochum, Indian Institute of Technology Kharagpur, Max Planck Institute for Security and Privacy, Delft University of Technology

AlphaXiv Trending

MemT-1.0: Towards Memory-Native Evolution
AlphaXiv
66

MemT-1.0: Towards Memory-Native Evolution

Views

Machine Learning
Text-to-Image Generative Modeling Without Language Pretraining
AlphaXiv
22

Text-to-Image Generative Modeling Without Language Pretraining

Views

Affiliation: Stanford University, Affiliation: Stanford University Affiliation: Radical Numerics, Stanford University

SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model
AlphaXiv
20

SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model

Views

NLP
Semantically Steerable Image Tokenization
AlphaXiv
16

Semantically Steerable Image Tokenization

Affiliation: QUVA Lab, University of Amsterdam, Affiliation: VIS Lab, University of Amsterdam, Affiliation: QUVA Lab, University of Amsterdam Affiliation: VIS Lab, University of Amsterdam, QUVA Lab, University of Amsterdam, VIS Lab, University of Amsterdam

Learning the 4D World through Point Trajectory Forecasting
AlphaXiv
14

Learning the 4D World through Point Trajectory Forecasting

Views

Affiliation: Seoul National University, Affiliation: Physics Sim Lab, Affiliation: New York University, Affiliation: Seoul National University Affiliation: RLWRLD, Seoul National University

Sage-A1.5: From Long-Horizon Execution to Scientific Exploration with Capability Composition
AlphaXiv
13

Sage-A1.5: From Long-Horizon Execution to Scientific Exploration with Capability Composition

Views

Sage-A1.5 Team, Shanghai Artificial Intelligence Laboratory

HuggingFace Daily Papers

Machine Learning
Training the Whole Agent, Not Just the Model
HuggingFace
4

Training the Whole Agent, Not Just the Model

Lingxiao Tang, Wei Tao, Yiwen Guo, Lingfeng Bao

Tencent

Reinforcement Learning
Self-Credit: Credit Assignment for Long-Horizon Agentic RL via Intra-group Rescaling
HuggingFace
3

Self-Credit: Credit Assignment for Long-Horizon Agentic RL via Intra-group Rescaling

Fangcong Yin, Ming Yin, Tanvi Bhandarkar, Aashu Singh, Eryk Helenowski

Meta

NLP
On the Tip of the Tongue: First-Token Cueing as a Behavioral Probe of Recall Failures in LLMs
HuggingFace
2

On the Tip of the Tongue: First-Token Cueing as a Behavioral Probe of Recall Failures in LLMs

Jaewon Sok, Jewon Yeom, Seonghyeon Park, Jeongjae Park, Hwiyeong Lee

Seoul National University

Robotics
BiDex-100: A Comprehensive Benchmark for Bimanual Dexterous Manipulation Across 100+ Simulation Tasks
HuggingFace
1

BiDex-100: A Comprehensive Benchmark for Bimanual Dexterous Manipulation Across 100+ Simulation Tasks

Yong Wang, Kecheng Zheng, Xing Zhu, Chao Xu, Yujun Shen

Robbyant Research

Reasoning
Axiomatic World Modeling for Physics Reasoning
HuggingFace
1

Axiomatic World Modeling for Physics Reasoning

Xinye Yang, Zhenyang Liu, Yuxuan Wang, Yuanyuan Lei

Yijunshens
World Models with Configurable Perception
HuggingFace
1

World Models with Configurable Perception

Yijun Shen, Delong Chen, Xianming Hu, Kai Zhang, Pascale Fung

AI Research Papers — October 21, 2026 Edition | Agentic Brew